small bug fix for large mnl simulation re: specification of interaction terms

[mxndrwgrdnr]

The original code I wrote to accommodate a list of tables of interaction terms breaks if the interaction terms aren't a list of dataframes but rather a single dataframe, which we don't want. This PR simply fixes the large mnl run() method to account for the latter case.

[mxndrwgrdnr]

Not sure if this fix elicits its own version increase, perhaps could just get included in the next release if there's one in the works?

[coveralls]

Coverage decreased (-0.2%) to 91.928% when pulling aa2e9ed on large_mnl_intx_term_fix into 3324905 on master.

[smmaurer]

Great! It looks like the Travis failures are unrelated to your changes, but would you mind adding a commit here as described in issue #110 to see if it fixes the builds?

I'll be getting back to urbansim_templates work in July.. Happy to make a point release for this in the meantime if it would make your workflow easier, or if you think other folks might run into the same bug.

[smmaurer]

Oh, right, I guess we're already into the 0.2 dev series of version numbers. You could increment this and call it 0.2.dev7. And probably no release on pip/conda until 0.2 is finished, unless there's a good reason.

[mxndrwgrdnr]

Just triggered the builds with the requirements.txt fix. We'll see how it goes. I don't personally need the version update/point release, and I don't think anyone else will run into this issue. I just wanted to make sure the bug and the fix was documented for the next release. Is there an easy way to just let this PR camp out until the next release?

[smmaurer]

Sweet, that worked! Fine by me to leave this PR unmerged -- i'll make sure it gets included in the next release.

[mxndrwgrdnr]

👍

[bouzaghrane]

@smmaurer any updates on this? can we schedule some time to chat about it?